Background and Purpose
It is commonly acknowledged that children today are traveling less often by active modes and less often without adult supervision. Potential negative effects of these joint trends may include, for children, a decrease in physical activity level and loss of developmental benefits from independently navigating their neighborhood environment, and, for society as a whole, an increase in vehicle miles traveled and associated negative externalities.This research models the propensity of children aged 6-16 to walk or bike to parks and school without an adult chaperone. A rich set of potential correlates was available from linked household, parent, and child surveys.This analysis extends existing work on children’s active travel in several ways: 1) focus on travel without an adult, 2) inclusion of school and a non-school destinations, 3) separate walk and bike models, 4) consideration of both parent and child attitudes and perceived social norms, 5) explicit inclusion of household rules limiting walking or bicycling.

Objectives
This research seeks to understand the relative contributions of various child, parent, household, and neighborhood level factors to children’s independent active travel to the park and to school. Specific barriers and incentives are considered that may be effective policy levers to increase independent active travel, while controlling for socio-demographic characteristics of respondents.

Methods
Data are drawn from the initial phase of the Family Activity Study, a three-year longitudinal intervention study in Portland, Oregon that began in 2010. This analysis uses survey questionnaires taken by 281 households, including 428 children aged 6-16 and 416 adults. Binomial logit  models are specified for two parent-reported independent travel behaviors by children: walk and bike to parks at least once per month and walk and bike to school at least once per month. Models are estimated separately for biking and walking. Attributes include proximity, household socio-demographics, rules, norms, attitudes, built environment, children’s related travel behavior and adults’ past behavior, and residential location choice factors.  Special consideration is given to household rule effects and their predictors as well as differences between walk/bike and school/non-school choice situations.

Results
Increasing distance strongly discouraged independent walk and bike travel to the park and to school. Age was also strongly negatively correlated with all types of independent travel. Female children were less likely to travel to parks without an adult, but no gender bias was evident for school travel. Active travel to school increased the chances of biking and walking to the park. Older siblings had a strong positive effect for three of the four mode/destination combinations. Children in single parent households were less likely to walk independently, and children from higher income households were less likely to travel to school independently by either mode. Restrictive household rules had larger impacts on non-school than on school travel. Other kids' behavior and 11-16 year olds’ attitudes toward helmets were found to have relatively small but significant correlations with biking behavior.  Younger children’s safety perceptions and parents’ childhood active travel to school likewise had relatively small but significant correlations with independent walking. A number of land-use and built environment measures were tested but found to be insignificant predictors after controlling for other attributes.in terms of household rules, a single parent or stranger danger concern increased the odds of having a stay in sight rule. Younger children and girls were also more likely to face the stay at home rule. Parents that trusted neighbors and chose the location due in part to good walking conditions were less likely to have the rule. Traffic concerns increased the probability of a rule against biking in the street, as did mixed land uses. Households that selected a neighborhood for its good biking conditions for kids and parents who were confident bicyclists themselves decreased the odds that kids were banned from biking in the street.

Conclusions
Proximity, household socio-demographics, rules, norms, attitudes, children’s related travel behavior and adults’ past experience were significantly correlated with children’s independent active travel.  As measured, land-use and built environment factors were not significant predictors. Taken together, the results underscore the importance of treating walk and bike travel separately, studying non-school trips, understanding household rule formation, and incorporating variables at child, adult, household, and neighborhood level.

Support / Funding Source
The Family Activity Study and this paper were funded by the Active Living Research program of the Robert Wood Johnson Foundation.February 2010 - January 2013.